Skip to main content
Version: 3.0

About the CubeCOS version 3.0.0 Release

πŸš€ Overview​

CubeCOS 3.0 is the first open source release of CubeCOS, introducing major improvements to the management interface and user experience. Key enhancements include alerting services, UI-based service monitoring and repair tools, and strengthened security features. End users can now get started by directly downloading CubeCOS and get started with our new cloud platform.

For a full list of updates, refer to the CubeCOS changelog.

✨ New Features​

CubeCOS Management Interface​

The new CubeCOS 3.0.0 management adopts the new COSMOS UI experience.

  • The Cube COSMOS management interface style brings a vibrant and rounded interface to CubeCOS.
  • New dashboards to provide a quick and clear overview of the cluster and integrations.

Triggers (previously Notifications)​

  • Renamed notifications to triggers for more precise, condition-based alerting.
  • Triggers define the conditions and when administrators receive alerts via Slack or email based on specific events or thresholds.

Tuning and Tuning Management​

  • Introducing CubeCOS Cluster Tunings β€” administrators can now apply cluster-wide or node-specific configurations directly from the UI.
  • Configure and apply settings at the cluster or node level to customize storage, network, and operating system behavior.

Support File Management​

  • Streamlined support bundle generation through the new UI experience for easier troubleshooting and faster collaboration with Bigstack support.

Events and Analysis​

  • Added event visualization with filtering by severity and category to monitor current and historical cluster health and activity.
  • Integrated log analysis into the instance action log tab for faster root cause identification and streamlined issue diagnosis.

Cluster Health Management​

  • Cluster health checks and component repair are now available in the CubeCOS UI, allowing administrators to monitor and restore services with a single click.
  • Quicker issue identification and error tracing through the new health status page, which now displays historical status data for all core services.
  • Integrated core services repair and status checks into the new COSMOS based UI interface.

βš™οΈ Improvements​

General​

  • Provide a new appliance.login.greeing tuning to enable greeting messages to be show in admin CLI login and management interface login form.

hex_sdk​

  • Enhanced Hex operation performance
  • Modularized hex_sdk functions for easier usage and management

Events and Analysis​

  • More detailed events page with built-in filters and dashboards to help administrators identity issues faster.
  • Analyze event frequency by host, instance, or system to identify the most recurring event IDs.

Security​

  • Root account access will no longer lock after the initial setup period of 7 days.
  • Added password authentication hardening to MongoDB usage and access.

πŸ›  Fixes​

  • Fixed USB lock behavior: USB access is now unlocked by default on unconfigured CubeCOS clusters and will automatically lock after configuration.
  • Rancher Node Driver: With hex_LMI removed, the node driver endpoint has moved to a new endpoint served via httpd.

πŸ”„ Compatibility​

The following Cube related products are currently supported by CubeCOS 3.0.0.

  • CubeCMP 2.0RC5
    • Prerequisites: CubeCOS version 3.0 or later with AppFramework deployed
  • CubeVDI Driver
    • Prerequisites: CubeCOS version 3.0 or later and CubeVDI essentials

For more detail on the Cube platform services, visit the Bigstack contact page.

⚠️ Known issues​

  • When removing a node from a cluster, the node is successfully evicted but the OSDs associated with the removed node are not fully removed. Manual steps are required to properly remove the OSDs and release resources. (CubeCOS#101)

  • The installer might crash during installation. (CubeCOS#87)

  1. Reattempt login or installation process to restart and kick off installation.
  2. It may require multiple inputs before the installation process begins.
  • Rancher operations may encounter errors when upgrading to CubeCOS 3.0.0 from previous versions (requires AppFramework usage).
  1. Changes in the Cube node driver was made in v3.0.0 to better align with standard usage
  2. The Cube node driver address needs to be updated to reflect the new driver access address
  3. Navigate to Rancher management, enter Cluster Management -> Drivers -> Node Drivers -> Cube.
  4. Update the new Cube node driver URL to http://<cluster-vip>:8080/static/nodedrivers/docker-machine-driver-cube

  • Users may encounter errors when accessing the CubeCOS dashboard and the Ceph management dashboard simultaneously.

    • The CubeCOS dashboard and Ceph dashboard share a conflicting SAML token, which may result in loss of access when used concurrently.
    • Use incognito mode or separate browsers to avoid SAML token conflicts when accessing the CubeCOS and Ceph dashboards simultaneously.

  • Metrics in the Home β†’ Health view may be temporarily unavailable during node failure or VIP migration. (CubeCOS#148)

Feature deprecation and EOL​

Management Interface​

  • The Horizon management interface is being replaced by the new Skyline interface. Horizon can still be accessed at https://<cluster-vip>/horizon but will be fully removed in a future release.
Last updated on by Brian Su